In recent years, with the progress of information processing technologies, analysis engines which analyze various data have been developed. Various analysis engines have been developed: for example, an analysis engine which extracts a person wearing specific clothes from moving image data, an analysis engine which extracts a person with a specific face from moving image data, an analysis engine which generates location information that traces the flow line of a person from moving image data, and an analysis engine which generates text data from voice data.
Moreover, analysis systems which can acquire various analysis processing results from input data have also been developed by combining a plurality of analysis engines of the same type or different types. For example, a developed system executes analysis processing that is processing moving image data inputted from a camera in parallel or in series with a flow line extraction engine, a face extraction engine, an age discrimination engine and so on and judging a person who behaves in a given manner. Such an analysis system that judges a person from moving image data captured by a camera is required to acquire an analysis processing result quickly without delay, preferably, in real time.
Patent Document 1: Japanese Unexamined Patent Application Publication (Translation of PCT Application) No. 2004-535616
However, in a case where analysis target data is data which is consecutively inputted and is comparatively large-volume, such as moving image data or voice data, the load of the analysis processing executed by the analysis engines is high, and it is difficult to quickly acquire the result of the analysis processing. On the contrary, in a case where the volume of analysis target data is made to be small in advance by, for example, setting the frame rate of moving image data low, there is a problem that an analysis result with desired analysis accuracy cannot be acquired.
Patent Document 1 discloses a technique that is changing pixels to be analyzed in accordance with the intrusion status of a target object into a range monitored by the camera. This technique is beneficial in a case where a desired analysis result can be acquired even if a region to be subjected to analysis processing is limited. However, this technique cannot work in other cases. Therefore, there is a problem that a desired analysis result cannot be acquired quickly.